Design and Development of a Short-Term Photovoltaic Power Output Forecasting Method Based on Random Forest, Deep Neural Network and LSTM Using Readily Available Weather Features.
Denis RangelovMichell BoergerNikolay TcholtchevPhilipp LämmelManfred HauswirthPublished in: IEEE Access (2023)
Keyphrases
- short term
- long term
- feature set
- random forest
- neural network
- load forecasting
- forecasting model
- arima model
- feature vectors
- chaotic time series
- decision trees
- feature space
- data sets
- genetic algorithm
- prediction model
- medium term
- moving average
- neural network model
- optimization method
- class labels
- training samples
- classification accuracy
- artificial neural networks
- similarity measure
- feature selection